GFCI and an Open Ground

road-runner wrote:

I think what's behind all the downstream grounded neutral debate is that when somebody diligently goes to the web to see how a GFCI works, they have a 90+% chance of pulling up an explanation that totally omits that grounded neutral detection even exists. And for the <10% that learn that it exists, they'll be led to believe that there's exactly one way that it's done, which is the trap I fell in to.

I was surprised this morning when doing some googling to find some decent info. to reference, how little there actually was - as in next to nothing. UL has also added additional requirements in recent years. If anyone has seen a straightforward explanation on how all the latest features/requirements work (like end of life and mis-wire for ex.), it would nice to see a link.

myredracer wrote:
Here's a couple of diagrams that show how the 2nd coil detects a grounded neutral current flow.

It's best for any of these GFCI info sources to not make broad specific statements on how the grounded neutral detection works. Better to present things as an example of how it can be done. As wnjj pointed out after digging into spec sheets, in at least a couple of the implementations the 2nd coil does not do any detecting, and the separate transformer has given way to an added coil on the current detecting transformer. Maybe there are one or more cases where the 2nd coil really does detect, so I'm trying to be cautious. Continuing the trend, I just found a chip that needs only a single coil winding for everything. I didn't see any explanation of how this one works. http://www.icbase.com/File/PDF/ONS/ONS54501409.pdf

I think what's behind all the downstream grounded neutral debate is that when somebody diligently goes to the web to see how a GFCI works, they have a 90+% chance of pulling up an explanation that totally omits that grounded neutral detection even exists. And for the <10% that learn that it exists, they'll be led to believe that there's exactly one way that it's done, which is the trap I fell in to.

Here's a couple of diagrams that show how the 2nd coil detects a grounded neutral current flow. If a path exists between neutral & ground after the GFCI, it completes a loop along with the normal neutral/ground connection at the main service panel (or sometimes the CG pedestal) and enough current flows to trip the GFCI. AFAIK, the grounded neutral coil has been required for at least a couple of decades. A load doesn't need to be connected to detect a grounded neutral and can sometimes cause confusion when trying to diagnose why a GFCI has tripped in an RV.

BurbMan wrote:

BB_TX wrote:
OK, my head is starting to hurt. After all the reading I still cannot grasp how the circuit can detect a fault between the neutral and an open (not connected) ground. So I will just accept that it happens. I was a systems engineer with little experience at the circuit level, so that may be my weakness here.

You are thinking about this all wrong BB....

The GFCI is designed to detect an imbalance in the current from the hot to the neutral. Even without a ground connection, an imbalance is created when YOU become the ground, ie you are getting shocked!

In the example given by the OP, the lack of a ground connection to the trailer will not trip the GFCI since it is monitoring the hot/neutral balance. Because the trailer is not grounded, if any part of the skin becomes electrified, you will get a shock when you touch it, because YOU become the ground path from the skin to the ground. In that case the GFCI will see that current is leaking from hot to ground THROUGH YOU and trip the circuit.

Burbman, your description is completely correct, however BB_TX was asking about the other aspect of GFCI which can detect a short between downstream neutral and ground with no loads or humans involved in the circuit. It does that by exploiting the concept called a ground loop. With the downstream short and the neutral/ground bonding in the main service panel, there are now 2 places where neutral and ground are connected. This forms a (long) loop of wire. The GFCI wraps a primary coil around the neutral and essentially uses that loop as a secondary coil to form a transformer. If current flows in that loop, it is detected by the original detector circuit as current in the neutral that is not in the hot.

BB_TX wrote:
OK, my head is starting to hurt. After all the reading I still cannot grasp how the circuit can detect a fault between the neutral and an open (not connected) ground. So I will just accept that it happens. I was a systems engineer with little experience at the circuit level, so that may be my weakness here.

You are thinking about this all wrong BB....

The GFCI is designed to detect an imbalance in the current from the hot to the neutral. Even without a ground connection, an imbalance is created when YOU become the ground, ie you are getting shocked!

In the example given by the OP, the lack of a ground connection to the trailer will not trip the GFCI since it is monitoring the hot/neutral balance. Because the trailer is not grounded, if any part of the skin becomes electrified, you will get a shock when you touch it, because YOU become the ground path from the skin to the ground. In that case the GFCI will see that current is leaking from hot to ground THROUGH YOU and trip the circuit.

If an RV Chassis is allowed to float while it is serviced by a 30 Amp shore power the RV chassis will be some where "less then" 120 / 2 = 60 volts CHASSIS to EARTH by virtue of stray capacitance between H and G and N and G in the RV.

Mex's No Fault

GLOVES

ktmrfs wrote:

also, remember that a GFCI will NOT provide 100% protection against electricution. If you are across hot and neutral AND insulated well enought from any ground path, a GFCI won't trip. Example. in the trailer no stab jacks down, dry ground, rubber soled shoes and you get across hot and neutral. Good chance there is not a good enough ground path to cause a GFCI to trip.

Same thing in a house. rubber shoes on dry hardwood or vinyl floor and you get across hot and neutral. good luck.

That's right, this fault already exists in every system, so it's not a net-new risk when the RV is floated relative to shore supply.

I have a scenario where I am feeling virtually compelled to float the RV from shore, and my plan is to apply a GFI/RCD to mitigate the new risk this introduces. That's why I'm posing this hypothetical. If the GFI truly mitigates the only meaningful new risk, then I'm good to go. If not, I need to reevaluate.

Harvard wrote:
Assuming there are now two faults, one being the floating chassis and the second being the person standing on EARTH being exposed to the 120VAC supply to the RV. Yes, the GFCI would trip.

Three faults, I suppose. The lack of an EGC, the hot being exposed somehow (hot to casing, etc.), and the person having low impedance to earth and also touching the exposed hot.

But if there is only one fault, the open ground (floating chassis), a person receiving the "tingle" while providing the path from the 40VAC CHASSIS to EARTH.
No, I do not think the GFCI would trip.

This I do not understand. How did we get a chassis to 40VAC? Separately, if the chassis is energized at any voltage, how is this different than the scenario already described above?

Thanks!

Harvard wrote:

nebster wrote:
So here's a follow-up question to confirm something I have been presuming for a while:

If the RV is powered via a line+neutral shore cord, with no EGC, and an inline GFI is provided in the shore cord... is a hot chassis event with a human inadvertently providing a low impedance path to earth ground something that the GFI will (should) protect against?

Assuming there are now two faults, one being the floating chassis and the second being the person standing on EARTH being exposed to the 120VAC supply to the RV. Yes, the GFCI would trip.

But if there is only one fault, the open ground (floating chassis), a person receiving the "tingle" while providing the path from the 40VAC CHASSIS to EARTH.
No, I do not think the GFCI would trip.

also, remember that a GFCI will NOT provide 100% protection against electricution. If you are across hot and neutral AND insulated well enought from any ground path, a GFCI won't trip. Example. in the trailer no stab jacks down, dry ground, rubber soled shoes and you get across hot and neutral. Good chance there is not a good enough ground path to cause a GFCI to trip.

Same thing in a house. rubber shoes on dry hardwood or vinyl floor and you get across hot and neutral. good luck.